My immediate reaction was that this was something that came at the cost of strategic decision. If an entire area consists of nothing but balanced encounters (yielding level-appropriate treasures), then there are really no encounter-selection decisions to be made. Whatever you attack, it will give you about the same challenge, scaled to your level, with roughly the same reward. There's no need to exercise selectivity in deciding whether or not to bash down a door and kill what's inside. It's just guaranteed to be something you can defeat, given average luck. Nor does it help to avoid that encounter, since the very next one will be about the same difficulty -- and yield about the same treasure. You'll just waste the time explaining how to avoid the encounter only to move on to a reskinned version of the exact same thing. So the game becomes a methodical room-clearing exercise, like many computer RPGs.
This isn't always necessarily a bad thing. In 4th edition D&D, in particular, the intent of WotC was clearly to make a good tactical miniature man-to-man combat system, and to allow that tactical combat to have a lot of richness and complexity. For someone wanting to luxuriate in a detailed tactical simulator with lots of crunchy gamist decisions to make, this is exactly the mechanics you want. Anything else would be silly -- like having everyone set up a Warhammer 40K Table, only to roleplay a parley between the Space Marines and the Tau that resulted in a truce that required everyone to pack up the table and go home.
In my current campaign, though, the ACKS rules pretty explicitly emphasize strategic-level decision. Character classes are designed around non-combat capabilities; a bard, say, is a wimpy fighter with powerful non-combat proficiency options. But more importantly, the game itself employs an old-school approach toward the integration of "balance" into the game's social contract. The philosophical contrast is this:
- In a modern RPG, balance is a one-sided GM responsibility.
- In a retro RPG, balance is a negotiated aspect of emergent play.
In an earlier post, I noted how old-school RPG play really has four distinct modes: (1) dungeon-crawling, (2) hex-crawling, (3) mass combat, and (4) strategic resource management. The "dungeon crawling game" bears a strong superficial resemblance to a modern packaged adventure path, with monsters populating a traditional dungeon. But the mechanics of the old-school dungeon-crawl are vastly different than in more modern systems because of the pivot away from session-long resource management. Really, most modern games have more in common with the "hex crawling game" in terms of structure, as isolated encounters that don't share resource pools.
The critical difference in the system is how hit points are understood. A modern game (4th edition onward, or any recent computer game) tends to interpret hit points as "luck", or "stamina", or "the favor of the gods". When they run out, the next hit kills you because you exhausted your supply of those intangibles. It's easy to regain those intangibles between battles with a "healing surge", or a "quick rest", or some similar mechanic. Even some spell-like effects work the same way, implemented as "per encounter" or "at will" abilities, or by allowing a quick rest to recover some or all spell slots (points, whatever). Death is replaced by a temporary knock-out effect. As a result, each combat can be balanced in roughly the same way: You assume everyone starts at full health, and build a mirror-image collection of enemies. This makes it perfectly sensible to put together a series of encounters that all have identical challenge ratings, maybe with a slightly harder boss fight at the end for drama.
A simulationist retro-RPG like ACKS extends resource management over the entire session. Hit points (understood as "flesh wounds" or "blood loss") can't easily be returned, since healing spells are pretty weak and natural healing takes forever. Spells are "per day", not "per encounter". Death is permanent, and serious injuries are realistically incapacitating. This means that putting together encounters that are "balanced" will beat down the party quickly (which might be fun for a Tomb of Horrors-esqe tournament module, but not for a long campaign). Instead, the responsibility of the GM is to ensure that most encounters are weak, and the few strong encounters are either avoidable or else are capstone events for which the party will be able to carefully plan and prepare (to wring every advantage out of combat-as-war strategems). The weak encounters never yield good rewards and drain resources, so it's more efficient to bypass them when possible. If you want a tough encounter with good rewards, you need to do the legwork to find one.
(Incidentally, I'd say that 3rd edition D&D and PFRPG are both centrist systems that encourage a little of both approaches.)
The party bears a much heavier level of responsibility for finding methods of identifying and classifying potentially difficult encounters, to know when to avoid them or how to prepare for them. Kicking down every door will eventually lead to a total party kill. Several common methods of gathering strategic information include:
- Dungeon geometry tropes: E.g. the deep caves have nastier monsters, so dwarves get a "detect slopes and depths underground" racial perk
- Scouting techniques: Listening at doors, for example, is usually prudent before bashing them down
- Parley with dungeon denizens
- Interrogation: Leave an orc alive after combat to see if you can get a sense of where the Big Bads are holed up (hopefully you took "orc" as a language!)
- Exploiting dungeon faction rivalries
- Disguise and infiltration
- Divination magic (once you figure out the right questions to ask!)
- When all else fails, exploring recklessly and then running away from that demilich...
The social contract for a GM is to provide a lot of non-fatal encounters that slowly nickel-and-dime the party's resources (spells and hit points) to the point of feeling progressively more weak and vulnerable as they approach a major objective -- so that "Do we really want to keep going?" is a tough call, not a no-brainer.